Every business idea is a stack of assumptions disguised as a plan. The plan says “I’ll build X for Y people who will pay Z.” But behind that plan are dozens of untested beliefs, any one of which could be wrong enough to kill the business.
I learned this painfully in 2018. My plan was sound: build a workshop series for startup founders on lean validation methods. The assumptions behind it — that founders would pay €200+ for workshops, that they’d attend in person in Graz, that they’d find me through LinkedIn posts — seemed reasonable. I didn’t question them because they felt true.
Two of three assumptions were wrong. Founders would pay — but not for in-person events in Graz (the audience was too small locally). And LinkedIn posts didn’t drive registrations (direct outreach did). The business eventually worked, but only after I identified and corrected the wrong assumptions. Those corrections cost me four months.
If I’d started by listing my assumptions, ranking them by risk, and testing the riskiest ones first, I would have discovered the problems in weeks instead of months.
What Assumptions Look Like
Every business idea contains assumptions in five categories.
Problem assumptions: “This problem is real, painful, and frequent enough to pay to solve.”
This is the foundation. If the problem isn’t real or isn’t painful enough, nothing else matters. Problem assumptions are often the easiest to test (through customer conversations) and the most commonly skipped.
Customer assumptions: “I know who has this problem, and I can find and reach them.”
You might be right about the problem but wrong about who has it. Or right about who has it but unable to reach them affordably. Customer assumptions determine whether you can build a sustainable acquisition engine.
Solution assumptions: “My proposed solution actually solves the problem in a way that people will prefer over their current approach.”
A real problem with a wrong solution is still a failed business. The gap between “this problem exists” and “my solution is the right answer” is where most product failures live.
Business model assumptions: “People will pay [price], I can deliver at [cost], and the math works.”
The unit economics of your business are assumptions until you test them. Price sensitivity, delivery costs, and margin calculations are all guesses until real transactions happen.
Channel assumptions: “I can reach my customers through [specific channel] at an affordable cost.”
Even a perfect product for a real problem fails if you can’t reach the customers affordably. Channel assumptions are the most commonly untested category.
Building Your Assumption Stack
Here’s the exercise I do with every new business idea.
Step 1: List every assumption. Write down everything that must be true for the business to work. Don’t filter. Don’t dismiss anything as “obvious.” If it’s something you believe but haven’t proven, it goes on the list.
A typical list has 15-25 assumptions. That feels like a lot, but most ideas are built on more assumptions than founders realize.
Step 2: Rank by impact. For each assumption, ask: “If this is wrong, does the business survive?” Score each assumption 1-5 based on the severity of being wrong. A “5” means the business dies if this assumption is wrong. A “1” means it’s suboptimal but recoverable.
Step 3: Rank by uncertainty. For each assumption, ask: “How confident am I that this is true?” Score 1-5 based on uncertainty. A “5” means you’re guessing. A “1” means you have strong evidence.
Step 4: Calculate risk priority. Risk = Impact × Uncertainty. The assumptions with the highest risk scores are the ones to test first.
Step 5: Design a test for each high-risk assumption.
For the top 3-5 assumptions by risk score, design a specific, fast, cheap test. “I’ll test the problem assumption by conducting 10 customer interviews this week.” “I’ll test the price assumption by pre-selling at the proposed price.” “I’ll test the channel assumption by running a small test campaign.”
A Real Assumption Stack Example
Here’s an assumption stack from a SaaS idea I evaluated last year.
| Assumption | Impact (1-5) | Uncertainty (1-5) | Risk Score |
|---|---|---|---|
| SME owners will pay for compliance tracking | 5 | 3 | 15 |
| They’ll pay €49/month specifically | 4 | 4 | 16 |
| We can reach them through LinkedIn | 3 | 4 | 12 |
| The tool can be built with no-code | 2 | 2 | 4 |
| Word of mouth will drive 30% of growth | 3 | 5 | 15 |
| Churn will be below 5% monthly | 4 | 5 | 20 |
| One founder can operate this solo | 3 | 3 | 9 |
The highest risk assumption: churn below 5% (score 20). This was the most uncertain and would kill the business if wrong. Testing it required getting 20+ customers and measuring actual retention — which meant the churn assumption couldn’t be tested until after other assumptions (problem, price, channel) were validated.
The second-highest: price (score 16). Testable immediately through pre-selling.
The testing order: price first (can test this week), problem (in parallel through customer conversations), channel (next week with a small LinkedIn campaign), then churn (after 1-2 months of real customers).
By testing in this order, we discovered the price was right but the channel was wrong — LinkedIn didn’t convert for this audience. We pivoted to industry-specific forums and the channel started working. If we’d assumed LinkedIn worked and built the whole marketing strategy around it, we’d have wasted months.
Testing Assumptions Cheaply
Each assumption type has a standard cheap test.
Problem assumption: 10-15 customer conversations. Cost: your time. Timeline: 1-2 weeks.
Customer assumption: Community research + direct outreach to 50 target profiles. Cost: your time. Timeline: 1 week.
Solution assumption: Wizard of Oz MVP or manual delivery to 5-10 customers. Cost: your time + maybe €50 in tools. Timeline: 2-4 weeks.
Price assumption: Pre-sell at proposed price to 40-50 targets. Cost: Stripe setup fee. Timeline: 1-2 weeks.
Channel assumption: Small test campaign on proposed channel with €100 budget or 50 organic posts. Cost: €100 or time. Timeline: 2-4 weeks.
The total cost of testing all five assumption types: under €200 and 6-8 weeks. Compare that to building a product for six months and discovering the assumptions were wrong. The assumption stack turns a €50,000 gamble into a €200 experiment.
When Assumptions Change
Assumptions aren’t static. As you learn from testing, assumptions get confirmed, disproven, or — most often — refined.
“SME owners will pay for compliance tracking” might refine into “SME owners in healthcare will pay for GDPR compliance tracking.” The broad assumption was sort-of-true, but the refined version is much more actionable. This refinement is the normal output of assumption testing — not a clean yes/no, but a sharper understanding of the conditions under which the assumption holds.
When a high-risk assumption is disproven, you face the pivot-or-persevere decision. The assumption stack makes this decision clearer because you know exactly which assumption failed and can evaluate whether it’s fixable (change the price, change the channel) or fundamental (the problem isn’t real enough).
Key Takeaways
- Every business idea is a stack of assumptions. List them all — typically 15-25 for a new idea.
- Rank assumptions by Risk = Impact x Uncertainty. Test the highest-risk assumptions first.
- Five categories of assumptions: problem, customer, solution, business model, and channel. Each can be tested cheaply within 1-4 weeks.
- Testing all high-risk assumptions costs under €200 and 6-8 weeks. This turns a months-long gamble into a structured experiment.
- Assumptions refine, not just confirm or deny. The output of testing is usually a sharper version of the original assumption, not a binary yes/no.
